Draft. These draft recommendations are not final and therefore are not intended for use or citation.

Transcription

1 ASH Recommendations for VTE in Non-Surgical Patients INTRODUCTION American Society of Hematology (ASH) guidelines are based on a systematic review of available evidence. Through a structured process, a guideline panel makes judgements about the evidence and forms recommendations. The public comment period occurs after recommendations are formed but before a manuscript report of the guidelines has been finalized and before ASH organizational approval of the guidelines. Comments collected during the open comment period are provided to the guideline panel for review prior to finalizing the guidelines. These draft recommendations are not final and therefore are not intended for use or citation. To submit comments on the draft recommendations, please visit Only comments submitted via the online survey will be reviewed by the guideline panel. The public comment period for these draft recommendations is July 24 August 25, RECOMMENDATIONS Question 1: Should heparin vs. no heparin be used in acutely ill medical patients for VTE prophylaxis? Question 2a and 2b: If heparin is recommended, should low molecular weight heparin (LMWH) vs. unfractionated heparin (UFH) be used in acutely ill medical patients? Question 3: If anticoagulants are recommended, should fondaparinux versus LMWH or UFH be used in acutely ill medical patients? The ASH guideline panel suggests using LMWH, fondaparinux or UFH rather than no heparin in acutely ill medical patients (conditional recommendation, low certainty of the evidence about effects). However, the ASH guideline panel suggests using LMWH (low certainty of the evidence about effects) or fondaparinux (very low certainty of the evidence about effects) rather than UFH in acutely ill medical patients (conditional recommendation). This recommendation applies to patients with stroke who receive VTE prophylaxis. Question 4: Should any DOAC vs. prophylactic LMWH be used in acutely ill medical patients? The ASH guideline panel recommends using prophylactic LMWH over DOACs in acutely ill medical inpatients (strong recommendation, moderate certainty of the evidence). Question 5: Should extended DOAC vs. shorter duration non-doac inpatient prophylaxis be used in acutely ill medical patients? The ASH guideline panel recommends not using extended DOACs (rather than shorter) inpatient prophylaxis with a non-doac agent in acutely ill medical patients (strong recommendation, moderate certainty of the evidence about effects). Question 6: Should extended duration (i.e., up to 30 or 40 days) vs. in hospital only heparin be used for the thromboprophylaxis of VTE in acutely ill hospitalized medical patients? The ASH guideline panel recommends against extended prophylaxis compared with shorter prophylaxis in acutely or critically ill medical patients (strong recommendation based on moderate certainty in the evidence about effects). Question 7: Should mechanical prophylaxis vs. no prophylaxis be used for medical patients (acutely and critically ill)? The ASH guideline panel suggests using mechanical prophylaxis over no prophylaxis in acutely and critically ill medical patients (conditional recommendation, moderate certainty of the evidence about effects).

2 Question 8: Should mechanical combined with pharmacological vs. mechanical prophylaxis alone be used for medical patients (acutely and critically ill)? The ASH guideline panel suggests using mechanical prophylaxis over mechanical combined with pharmacological prophylaxis in acutely and critically ill medical patients (conditional recommendation, very low certainty of the evidence about effects). Question 9: Should mechanical combined with pharmacological vs. pharmacological prophylaxis alone be used for medical patients (acutely and critically ill)? The ASH guideline panel suggests pharmacological prophylaxis or mechanical alone over mechanical combined with pharmacological in acutely and critically ill medical patients (conditional recommendation, very low certainty about effects). Question 10: Should mechanical prophylaxis vs. pharmacological prophylaxis be used for medical patients (acutely and critically ill)? The ASH guideline panel suggests using pharmacological prophylaxis over mechanical prophylaxis in acutely or critically ill medical patients (conditional recommendation, very low certainty of the evidence about effects). Question 11: Should pneumatic compression devices vs. graduated compression stockings be used for medical patients (acutely and critically ill)? The ASH guideline panel suggests using either pneumatic compression devices or graduated compression stockings in acutely and critically ill medical patients (conditional recommendation, very low certainty in the evidence about effects). Question 12: Should heparin vs. no heparin be used in critically ill patients? The ASH guideline panel recommends using heparin over no heparin in critically ill patients (strong recommendation based on moderate certainty in the evidence about effects). Question 13: Should low molecular weight heparin (LMWH) vs. unfractionated heparin (UFH) be used in critically ill patients? The ASH guideline panel suggests using LMWH over UFH in critically ill patients (conditional recommendation based on moderate certainty of the evidence about effects). Question 14: Should thromboprophylaxis vs. no thromboprophylaxis be used in chronically ill medical inpatients (including nursing home patients)? The ASH guideline panel suggests not using thromboprophylaxis (compared to using thromboprophylaxis) in chronically ill medical inpatients including nursing home patients (conditional recommendation based on very low certainty in the evidence about effects). Question 15: Should LMWH vs. no LMWH be used for long distance (>4 hours) travelers? Question 16: Should aspirin vs. no asprin be used for long distance (>4 hours) travelers? Question 17: Should graduated compression stockings (GCS) vs. no GCS be used for long distance (>4 hours) travelers? The ASH guideline panel suggests not using graduated compression stockings for long distance (> 4 hours) travel in people without known risk factors (conditional recommendation, very low certainty in the evidence about effects). The ASH guideline panel suggests against using LMWH in long-distance (>4 hours) travelers without known risk factors for VTE (conditional recommendation, very low certainty of the evidence about effects). The ASH guideline panel suggests against using Aspirin in long-distance (>4 hours) travelers without known risk factors for VTE (conditional recommendation, very low certainty of the evidence about effects). People without known risk factors who place a high value on prevention of VTE, may choose using GCS (also reduces edema). In people who are at substantially increased VTE risk (e.g. recent surgery, prior history of VTE, hormone replacement therapy, pregnant or postpartum women, active malignancy or two or more risk factors) the ASH guideline panel suggests using graduated compression stockings or prophylactic LMWH for long distance (> 4 hours) travel (conditional recommendation, very low certainty in the effects). evidence about

3 In people who are at substantially increased VTE risk (e.g. recent surgery, prior history of VTE, hormone replacement therapy, pregnant or postpartum women, active malignancy or two or more risk factors) and where LMWH or GCS is not feasible (e.g. resource constrained setting or aversion to other anticoagulants), the ASH VTE guideline committee suggests using aspirin rather than no treatment (conditional recommendation, very low certainty in the evidence about effects). Question 18: Should thromboprophylaxis vs. no thromboprophylaxis be used in medical outpatients with minor provoking factors for VTE (immobility, minor injury, outpatient illness/infection)? The ASH guideline panel suggests not using thromboprophylaxis in medical outpatients patients with minor provoking factors for VTE (immobility, minor injury, outpatient illness/infection) (conditional recommendation, very low certainty of the evidence).

4 Question 1 Should heparin vs. no heparin be used for acutely ill medical patients? POPULATION: acutely ill medical patients BACKGROUND: Hospitalization is a major risk factor for occurrence of thrombosis in acutely ill medical patients. Studies have shown an increased INTERVENTION: heparin risk of venous thromboembolism in the first 30 days after patients discharge. Venous thromboembolism is one of the most significant COMPARISON: no heparin preventable causes of morbidity and mortality in hospitalized patients. MAIN OUTCOMES: SETTING: PERSPECTIVE: Assessment PROBLEM DESIR ABLE Mortality; Pulmonary Embolism - representing the moderate marker state; Proximal Deep Vein Thrombosis - representing the moderate marker state; Distal Deep Vein Thrombosis - representing the moderate marker state; Major Bleeding; Gastrointestinal Bleeding; Heparin-Induced Thrombocytopenia; Inpatient Clinical recommendation - population perspective JUDGEMENT RESEARCH EVIDENCE ADDITIONAL CONSIDERATIONS Is the problem a priority? No Probably no Probably yes Yes How substantial are the desirable anticipated effects? Hospitalization represents a major risk factor for deep vein thrombosis and pulmonary embolism; up to 22% of venous thromboembolisms occur following non-surgical hospital admissions (Heit 2002). Symptomatic venous thromboembolism may complicate over 4.5 per 1000 admissions to the medical services (Zakai 2013), leading to a higher risk of morbidity and mortality. See Appendix 1

5 UNDESIRABLE EFFECTS CERTAINTY OF EVIDENCE VALUES Trivial Small Moderate Large How substantial are the undesirable anticipated effects? Large Moderate Small Trivial What is the overall certainty of the evidence of effects? Very low Low Moderate High No included studies Is there important uncertainty about or variability in how much people value the main outcomes? Important uncertainty or variability Possibly important uncertainty or variability Probably no important uncertainty or variability No important uncertainty or variability Additional evidence for PICC lines as a risk factor: Greene (2015) reported, after risk adjustment, PICC use was independently associated with all-cause VTE ([HR], 3.16; 95% [CI], ), upperextremity DVT (HR, 10.49; 95% CI, ), and lower-extremity DVT (HR, 1.48; 95% CI, ). PICC use was not associated with PE (HR, 1.34; 95% CI, ). Evans (2013) reported on a population with a mix of medical conditions where the ORs for DVT associated with PICC diameter were: o Double-lumen 5F vs single-lumen 4F: OR= 2.24 o Triple-lumen 6F vs single-lumen 4F: OR= 6.35 o Triple-lumen 6F vs double-lumen 5F: OR= 2.84 o Triple-lumen 5F vs single-lumen 4F: OR= 0.49 o Triple-lumen 5F vs double-lumen 5F: OR= 0.22 o Triple-lumen 5F vs triple-lumen 6F: OR= The relative importance of the outcomes* was as follows in the identified studies: Pulmonary embolism: (Hogg 2013, Hogg 2014, Locadia 2004) Deep vein thrombosis: (Hogg 2013, Hogg 2014, Locadia 2004, Marvig 2015, Utne 2016) Deep vein thrombosis patients' own current health: 0.95 (Time trade off) (Locadia 2004) Gastrointestinal tract bleeding event: 0.65 (standard gamble and time trade off) (Hogg 2013, Locadia 2004)

6 BALANCE OF EFFECTS Does the balance between desirable and undesirable effects favor the intervention or the comparison? Favors the comparison Probably favors the comparison Does not favor either the intervention or the comparison Probably favors the intervention Favors the intervention Muscular bleeding: 0.76 (time trade off) (Locadia 2004) Minor intracranial bleeding event: 0.75 (standard gamble) (Hogg 2013) Major intracranial bleeding event: 0.15 (standard gamble) (Hogg 2013) Central nervous system bleeding: (standard gamble) (Lenert 1997, O'Meara 1994) Treatment with LMWH: (time trade off) (Marchetti 2001) * indicated by utility value where 0 = death and 1.0 = full health Studies described the following regarding the relative importance of outcomes and patients' preferences for VTE prophylaxis; Patients highly value the benefits of VTE risk reduction of VTE prophylaxis (Haac 2016, Locadia 2004, Quante 2012, Wong 2015) and that they would like to avoid adverse events but most of them are not afraid of the adverse events (Barcellona 2000, Haac 2016, O Meara 1994, Quante 2012, Wong 2015). Patients highly value the benefits of VTE risk reduction of VTE prophylaxis; patients would like to avoid adverse events but most of them are not afraid of the adverse events. Studies additionally described the following regarding the relative importance of outcomes and patients preferences for the pharmacological prophylaxis: Most patients (78%) receiving low molecular weight heparin would like to continue with the same methods (Maxwell 2002).

7 RESOURCES REQUIRED How large are the resource requirements (costs)? Large costs Moderate costs Negligible costs and savings Moderate savings Large savings Cost of disease * See Appendix 2 * Unless otherwise specified, the costs of disease are based on a systematic review on the USA setting. (Grosse 2016) Treatment of an acute VTE on average appears to be associated with incremental direct medical costs of $12,000 to $15,000 (2014 US dollars) among first-year survivors, controlling for risk factors. Subsequent complications are conservatively estimated to increase cumulative costs to $18,000 23,000 per incident case. According to the analysis of Caprini et al. based on Medicare fee schedules, the predicted first-year treatment costs are $1100 for mild-moderate PTS (without leg ulcers) and $5000 for severe PTS with open ulcers. Cost estimates per bleed event vary widely. The average cost of heparin-induced thrombocytopenia (HIT) is $ Patients who develop CTEPH are much more expensive to treat than other PE patients, with average monthly expenditures of $5500 and $600, respectively. Cost of interventions See Appendix 3 Footnotes: ASP Drug Pricing Files: 2. National Average Drug Acquisition Cost, and The Affordable Care Act Federal Upper Limits (FUL): 3. Good Px: Only for medication costs, heparin is not costly. Low molecular weight heparins also increase cost, with enoxaparin (4 syringes 40mg/0.4ml of enoxaparin: $28.72 to $284.00), cost similar to heparin, and higher medication cost for dalteparin (10 syringes of dalteparin 18000iu/0.72ml: $1, to $1,576.44) and fondaparinux (10 syringes of fondaparinux 7.5mg/0.6ml: $4, $11,320.18). to

8 CERTAINTY OF EVIDENCE OF REQUIRED RESOURCES COST EFFECTIVENESS EQUITY What is the certainty of the evidence of resource requirements (costs)? Very low Low Moderate High No included studies Does the cost-effectiveness of the intervention favor the intervention or the comparison? Favors the comparison Probably favors the comparison Does not favor either the intervention or the comparison Probably favors the intervention Favors the intervention No included studies What would be the impact on health equity? Reduced Probably reduced Probably no impact Probably increased Increased Three reports (Lamy 2002, Offord 2003, Pechevis 2000) compared the costeffectiveness of heparin vs no heparin in medical patients, with all of them on enoxaparin. All of the studies suggested enoxaparin was favoured in the analyses and shown to be cost-effective. No research evidence was identified.

9 ACCEPTABILITY Is the intervention acceptable to key stakeholders? No Probably no Probably yes Yes Acceptability and use of pharmacological prophylaxis: Studies and surveys suggest the following regarding barriers associated with the intervention and its use: A survey among 568 physicians and 825 patients from 5 countries showed that more patients considered injectable treatments effective than considered oral treatments effective (87% versus 76%, respectively). This trend was well predicted by the physicians (98% and 61%, respectively). Additionally, 46% of patients would accept an injectable treatment program lasting >2 months (67% for life-threatening diseases), a figure underestimated by physicians (11% and 46%, respectively). Overall, 73% of patients stated they would never miss an injection, where as 54% of physicians expected patients to miss one injection in a month of therapy. (Cimminiello 2012) Among 250 hospitalized (surgical and medical) patients, initiation of prescribed therapy was 95% for LMWH, 88% for UFH 3/day and 87% for UFH 2/day. All scheduled doses were received by 77% on LMWH, 54% on UFH 3/day and 45% on UFH 2/day. Patient refusal explained 39% of omitted LMWH and 44% of omitted UFH doses. LMWH was less likely to be administered in surgical than in medical patients. (Fanikos 2010) A survey among 1,553 Canadian health care providers showed that DVT prophylaxis was perceived as important by all provider groups, but this did not appear to translate into knowledge about underutilization of current DVT prophylaxis strategies. Physicians and pharmacists recognized the underuse of DVT prophylaxis in medical patients, while nurses and physiotherapists tended to perceive prophylaxis strategies as appropriate. Lack of clear indications and contraindications for prophylaxis and concerns about bleeding risks were perceived as important barriers. Preprinted orders were considered the most potentially successful and feasible way to optimize prophylaxis. (Lloyd 2012) One large study using databases in the US found that the majority of at-risk hospitalized medically ill patients do not receive VTE prophylaxis. Only 18% of at-risk patients received VTE prophylaxis on day 1 or 2 in hospital, typically with LMWH (56% of patients receiving prophylaxis), pneumatic compression device (25%), vitamin K antagonist use (16%), or graduated compression stockings (11%). Use of prophylaxis exceeded 25% only in patients admitted from nursing homes and those with prior VTE. (Pendergraft 2013) Prescribing and uptake in different settings: Among 170 medical patients eligible for VTE prophylaxis, 54% received pharmacological VTE prophylaxis and 25% received non-pharmacological VTE prophylaxis due to a contraindication for pharmacological prophylaxis. (Panju 2011) Among 64 medical patients, 59% received appropriate VTE prophylaxis using LMWH. (Eijgenraam 2015)

10 Is the intervention feasible to implement? Feasibility and use of pharmacological prophylaxis: Studies showed the following barriers to utilizing the intervention/option: FEASIBILITY No Probably no Probably yes Yes Among 1,894 patients acutely ill medical patients from 29 Canadian hospitals, 23% received some form of VTE prophylaxis, but only 16% received appropriate prophylaxis. Factors independently associated with greater use of prophylaxis included internist (vs. other specialty) as attending physician, university-associated (vs. community) hospital, immobilization, presence of >1 VTE risk factors, and duration of hospitalization, however, use of prophylaxis was unacceptably low in all groups. (Kahn 2007) A survey among ICU directors, bedside pharmacists, thromboprophylaxis research coordinators and physician site investigators in 27 Canadian ICU s, showed that drug acquisition cost, fear of bleeding, lack of resident education, concern about renal failure, and habits were the top five barriers to LMWH use. Top five reported facilitators were preprinted orders, education, daily reminders, audit and feedback, and local quality improvement initiatives. Acceptability of facilitators varied across ICU s. (Cook 2014)

11 Summary of judgements JUDGEMENT IMPLICATIONS PROBLEM No Probably no Probably yes Yes Varies Don't know DESIRABLE EFFECTS Trivial Small Moderate Large Varies Don't know UNDESIRABLE EFFECTS CERTAINTY OF EVIDENCE VALUES BALANCE OF EFFECTS RESOURCES REQUIRED CERTAINTY OF EVIDENCE OF REQUIRED RESOURCES Large Moderate Small Trivial Varies Don't know Very low Low Moderate High Important uncertainty or variability Favors the comparison Large costs Possibly important uncertainty or variability Probably favors the comparison Moderate costs Probably no important uncertainty or variability Does not favor either the intervention or the comparison Negligible costs and savings No important uncertainty or variability Probably favors the intervention Very low Low Moderate High Favors the intervention Varies No included studies Don't know Moderate savings Large savings Varies Don't know No included studies COST EFFECTIVENESS Favors the comparison Probably favors the comparison Does not favor either the intervention or the comparison Probably favors the intervention Favors the intervention Varies No included studies

12 JUDGEMENT IMPLICATIONS EQUITY Reduced Probably reduced Probably no impact Probably increased Increased Varies Don't know ACCEPTABILITY No Probably no Probably yes Yes Varies Don't know FEASIBILITY No Probably no Probably yes Yes Varies Don't know Conclusions Should heparin vs. no heparin be used in acutely ill medical patients? TYPE OF RECOMMENDATION RECOMMENDATION Strong recommendation against the intervention Conditional recommendation against the intervention Conditional recommendation for either the intervention or the comparison Conditional recommendation for the intervention Strong recommendation for the intervention The ASH guideline panel suggests using LMWH, Fondaparinux or UFH rather than no heparin in acutely ill medical patients (conditional recommendation, low certainty of the evidence about effects). However, the ASH guideline panel suggests using LMWH (low certainty of the evidence about effects) or Fondaparinux (very low certainty of the evidence about effects) rather than UFH in acutely ill medical patients (conditional recommendation). This recommendation applies to patients with stroke who receive VTE prophylaxis. JUSTIFICATION SUBGROUP CONSIDERATIONS For patients with PICC and central line, this population is not considered separately in the trials and prominence of lines has increased since the trials have been done. UE DVT was excluded as an endpoint in the trials, and we could not find any data. IMPLEMENTATION CONSIDERATIONS

13 MONITORING AND EVALUATION RESEARCH PRIORITIES Research priorities include: Optimal baseline risk assessment. Optimal dosing of agents for prophylactic use. Role for heparins / LMWH and the dose in obesity and renal disease.

14 Appendices Appendix 1 Outcomes Mortality assessed with: all-cause mortality Pulmonary Embolism - representing the moderate marker state assessed with: symptomatic PE Proximal Deep Vein Thrombosis - representing the moderate marker state assessed with: symptomatic proximal DVT of participants (studies) Follow up (20 RCTs) (12 RCTs) 3706 (1 RCT) Quality of the evidence (GRADE) LOW a,b MODERATE b,c MODERATE a Relative effect (95% CI) RR 0.98 (0.92 to 1.04) RR 0.61 (0.46 to 0.81) RR 0.28 (0.06 to 1.37) Anticipated absolute effects * (95% CI) Risk with no heparin Study population Risk difference with heparin 69 per 1,000 1 fewer per 1,000 (6 fewer to 3 more) Study population 10 per 1,000 4 fewer per 1,000 (5 fewer to 2 fewer) Low 0 per 1,000 0 fewer per 1,000 (0 fewer to 0 fewer) Moderate 0 per 1,000 d 0 fewer per 1,000 (0 fewer to 0 fewer) Study population 4 per 1,000 3 fewer per 1,000 (4 fewer to 1 more) Low

15 Outcomes of participants (studies) Follow up Quality of the evidence (GRADE) Relative effect (95% CI) Anticipated absolute effects * (95% CI) Risk with no heparin Risk difference with heparin 0 per 1,000 0 fewer per 1,000 (0 fewer to 0 fewer) Distal Deep Vein Thrombosis - representing the moderate marker state assessed with: symptomatic distal DVT 3706 (1 RCT) Major Bleeding (15 RCTs) MODERATE a LOW a,b RR 0.75 (0.17 to 3.34) RR 1.48 (0.79 to 2.80) Moderate 1 per 1,000 d 1 fewer per 1,000 (1 fewer to 0 fewer) Study population 2 per 1,000 1 fewer per 1,000 (2 fewer to 5 more) Low 1 per 1,000 0 fewer per 1,000 (1 fewer to 2 more) Moderate 14 per 1,000 4 fewer per 1,000 (12 fewer to 34 more) Study population 7 per 1,000 3 more per 1,000 (1 fewer to 13 more) Gastrointestinal Bleeding 185 (2 RCTs) LOW a,b RR 2.61 (0.36 to 18.86) Study population 31 per 1, more per 1,000 (20 fewer to 558 more)

16 Outcomes of participants (studies) Follow up Quality of the evidence (GRADE) Relative effect (95% CI) Anticipated absolute effects * (95% CI) Risk with no heparin Risk difference with heparin Thrombocytopenia (3 RCTs) MODERATE a RR 0.95 (0.47 to 1.92) Study population 2 per 1,000 0 fewer per 1,000 (1 fewer to 2 more) a. Confidence interval includes appreciable benefit as well as no benefit, crossing the line of no effect. b. Downgraded for risk of bias due to unclear random sequence generation and allocation concealment across majority of trials, and lack of blinding of participants and study personnel as well as unclear blinding of outcome assessors. c. Not downgraded for imprecision. While confidence interval around relative effect is wide, the confidence interval around the absolute effect is narrow due to the small number of events and very large sample size. d. Data assumptions from Jones 2008

17 Appendix 2

18 Appendix 3

19 Question 2a & 2b Should LMWH vs. UFH be used for acutely ill medical patients? POPULATION: acutely ill medical patients BACKGROUND: Hospitalization is a major risk factor for occurrence of thrombosis in acutely ill medical patients. Studies have shown an increased INTERVENTION: LMWH risk of venous thromboembolism in the first 30 days after patients discharge. Venous thromboembolism is one of the most significant COMPARISON: UFH preventable causes of morbidity and mortality in hospitalized patients. MAIN OUTCOMES: SETTING: PERSPECTIVE: Assessment PROBLEM Mortality; Pulmonary Embolism - representing the moderate marker state; Proximal Deep Vein Thrombosis - representing the moderate marker state; Distal Deep Vein Thrombosis - representing the moderate marker state; Major Bleeding; Thrombocytopenia; Gastrointestinal Bleeding; Inpatient Clinical recommendation - population perspective Is the problem a priority? No Probably no Probably yes Yes JUDGEMENT RESEARCH EVIDENCE Hospitalization represents a major risk factor for deep vein thrombosis and pulmonary embolism; up to 22% of venous thromboembolisms occur following non-surgical hospital admissions (Heit 2002). Symptomatic venous thromboembolism may complicate over 4.5 per 1000 admissions to the medical services (Zakai 2013), leading to a higher risk of morbidity and mortality. ADDITIONAL CONSIDERATIONS DESIR ABLE How substantial are the desirable anticipated effects? See Appendix 1

20 Trivial Small Moderate Large UNDESIRABLE EFFECTS CERTAINTY OF EVIDENCE VALUES How substantial are the undesirable anticipated effects? Large Moderate Small Trivial What is the overall certainty of the evidence of effects? Very low Low Moderate High No included studies Is there important uncertainty about or variability in how much people value the main outcomes? Important uncertainty or variability Possibly important uncertainty or variability Probably no important uncertainty or variability No important uncertainty or variability The relative importance of the outcomes* was as follows in the identified studies: Pulmonary embolism: (Hogg 2013, Hogg 2014, Locadia 2004) Deep vein thrombosis: (Hogg 2013, Hogg 2014, Locadia 2004, Marvig 2015, Utne 2016) Deep vein thrombosis patients' own current health: 0.95 (Time trade off) (Locadia 2004) Gastrointestinal tract bleeding event: 0.65 (standard gamble and time trade off) (Hogg 2013, Locadia 2004)

21 BALANCE OF EFFECTS Does the balance between desirable and undesirable effects favor the intervention or the comparison? Favors the comparison Probably favors the comparison Does not favor either the intervention or the comparison Probably favors the intervention Favors the intervention Muscular bleeding: 0.76 (time trade off) (Locadia 2004) Minor intracranial bleeding event: 0.75 (standard gamble) (Hogg 2013) Major intracranial bleeding event: 0.15 (standard gamble) (Hogg 2013) Central nervous system bleeding: (standard gamble) (Lenert 1997, O'Meara 1994) Treatment with LMWH: (time trade off) (Marchetti 2001) * indicated by utility value where 0 = death and 1.0 = full health Studies described the following regarding the relative importance of outcomes and patients' preferences for VTE prophylaxis; Patients highly value the benefits of VTE risk reduction of VTE prophylaxis (Haac 2016, Locadia 2004, Quante 2012, Wong 2015) and that they would like to avoid adverse events but most of them are not afraid of the adverse events (Barcellona 2000, Haac 2016, O Meara 1994, Quante 2012, Wong 2015). Patients highly value the benefits of VTE risk reduction of VTE prophylaxis; patients would like to avoid adverse events but most of them are not afraid of the adverse events. Studies additionally described the following regarding the relative importance of outcomes and patients preferences for the pharmacological prophylaxis: Most patients (78%) receiving low molecular weight heparin would like to continue with the same methods (Maxwell 2002).

22 RESOURCES REQUIRED How large are the resource requirements (costs)? Large costs Moderate costs Negligible costs and savings Moderate savings Large savings Cost of disease * See Appendix 2 * Unless otherwise specified, the costs of disease are based on a systematic review on the USA setting. (Grosse 2016) Treatment of an acute VTE on average appears to be associated with incremental direct medical costs of $12,000 to $15,000 (2014 US dollars) among first-year survivors, controlling for risk factors. Subsequent complications are conservatively estimated to increase cumulative costs to $18,000 23,000 per incident case. According to the analysis of Caprini et al. based on Medicare fee schedules, the predicted first-year treatment costs are $1100 for mild-moderate PTS (without leg ulcers) and $5000 for severe PTS with open ulcers. Cost estimates per bleed event vary widely. The average cost of heparin-induced thrombocytopenia (HIT) is $ Patients who develop CTEPH are much more expensive to treat than other PE patients, with average monthly expenditures of $5500 and $600, respectively. Cost of interventions See Appendix 3 Footnotes: ASP Drug Pricing Files: Fee-for-Service-Part-B-Drugs/McrPartBDrugAvgSalesPrice/2016ASPFiles.html 2. National Average Drug Acquisition Cost, and The Affordable Care Act Federal Upper Limits (FUL): 3. Good Px: The panel considered that there are savings with LMWH. Only for medication costs, heparin is not costly. Low molecular weight heparins also increase cost, with enoxaparin (4 syringes 40mg/0.4ml of enoxaparin: $28.72 to $284.00), cost similar to heparin, and higher medication cost for dalteparin (10 syringes of dalteparin 18000iu/0.72ml: $1, to $1,576.44) and fondaparinux (10 syringes of fondaparinux 7.5mg/0.6ml: $4, to $11,320.18).

23 CERTAINTY OF EVIDENCE OF REQUIRED RESOURCES COST EFFECTIVENESS EQUITY What is the certainty of the evidence of resource requirements (costs)? Very low Low Moderate High No included studies Does the cost-effectiveness of the intervention favor the intervention or the comparison? Favors the comparison Probably favors the comparison Does not favor either the intervention or the comparison Probably favors the intervention Favors the intervention No included studies What would be the impact on health equity? Reduced Probably reduced Probably no impact Probably increased Increased Eleven reports (Amin 2009, Deitelzweig 2008, Fowler 2014, Leykum 2006, Lloyd 2001, McGarry 2004, McGarry 2006, Offord 2003, Schadlich 2006, Shorr 2007, Wilbur 2011) compared the cost-effectiveness of LMWH with UFH in hospitalized patients, one was a trial based analysis and another one was a decision-analytic model. All reports concluded LMWH was costeffective for thromboprophylaxis, and four of them showed that LMWH was more effective and provided net savings when compared to UFH (Amin 2009, Deitelzweig 2008, Offord 2003, Schadlich 2006). No research evidence was identified. Panel noted that costeffectiveness varies by country. Cost of UFH has increased recently.

24 ACCEPTABILITY Is the intervention acceptable to key stakeholders? No Probably no Probably yes Yes Acceptability and use of pharmacological prophylaxis: Studies and surveys suggest the following regarding barriers associated with the intervention/option and its use: A survey among 568 physicians and 825 patients from 5 countries showed that more patients considered injectable treatments effective than considered oral treatments effective (87% versus 76%, respectively). This trend was well predicted by the physicians (98% and 61%, respectively). Additionally, 46% of patients would accept an injectable treatment program lasting >2 months (67% for life-threatening diseases), a figure underestimated by physicians (11% and 46%, respectively). Overall, 73% of patients stated they would never miss an injection, where as 54% of physicians expected patients to miss one injection in a month of therapy. (Cimminiello 2012) Among 250 hospitalized (surgical and medical) patients, initiation of prescribed therapy was 95% for LMWH, 88% for UFH 3/day and 87% for UFH 2/day. All scheduled doses were received by 77% on LMWH, 54% on UFH 3/day and 45% on UFH 2/day. Patient refusal explained 39% of omitted LMWH and 44% of omitted UFH doses. LMWH was less likely to be administered in surgical than in medical patients. (Fanikos 2010) A survey among 1,553 Canadian health care providers showed that DVT prophylaxis was perceived as important by all provider groups, but this did not appear to translate into knowledge about underutilization of current DVT prophylaxis strategies. Physicians and pharmacists recognized the underuse of DVT prophylaxis in medical patients, while nurses and physiotherapists tended to perceive prophylaxis strategies as appropriate. Lack of clear indications and contraindications for prophylaxis and concerns about bleeding risks were perceived as important barriers. Preprinted orders were considered the most potentially successful and feasible way to optimize prophylaxis. (Lloyd 2012) One large study using databases in the US found that the majority of at-risk hospitalized medically ill patients do not receive VTE prophylaxis. Only 18% of at-risk patients received VTE prophylaxis on day 1 or 2 in hospital, typically with LMWH (56% of patients receiving prophylaxis), pneumatic compression device (25%), vitamin K antagonist use (16%), or graduated compression stockings (11%). Use of prophylaxis exceeded 25% only in patients admitted from nursing homes and those with prior VTE. (Pendergraft 2013) Panel noted that fewer injections make LMWH more acceptable. Prescribing and uptake in different settings: Among 170 medical patients eligible for VTE prophylaxis, 54% received pharmacological VTE prophylaxis and 25% received non-pharmacological VTE prophylaxis due to a contraindication for pharmacological prophylaxis. (Panju 2011) Among 64 medical patients, 59% received appropriate VTE prophylaxis using LMWH. (Eijgenraam 2015)

25 FEASIBILITY Is the intervention feasible to implement? No Probably no Probably yes Yes Summary of judgements Feasibility and use of pharmacological prophylaxis: Studies showed the following barriers to utilizing the intervention/option: Among 1,894 patients acutely ill medical patients from 29 Canadian hospitals, 23% received some form of VTE prophylaxis, but only 16% received appropriate prophylaxis. Factors independently associated with greater use of prophylaxis included internist (vs. other specialty) as attending physician, university-associated (vs. community) hospital, immobilization, presence of >1 VTE risk factors, and duration of hospitalization, however, use of prophylaxis was unacceptably low in all groups. (Kahn 2007) A survey among ICU directors, bedside pharmacists, thromboprophylaxis research coordinators and physician site investigators in 27 Canadian ICU s, showed that drug acquisition cost, fear of bleeding, lack of resident education, concern about renal failure, and habits were the top five barriers to LMWH use. Top five reported facilitators were preprinted orders, education, daily reminders, audit and feedback, and local quality improvement initiatives. Acceptability of facilitators varied across ICU s. (Cook 2014) JUDGEMENT PROBLEM No Probably no Probably yes Yes Varies Don't know DESIRABLE EFFECTS Trivial Small Moderate Large Varies Don't know UNDESIRABLE EFFECTS CERTAINTY OF EVIDENCE Large Moderate Small Trivial Varies Don't know Very low Low Moderate High No included studies IMPLICATIONS VALUES Important uncertainty or variability Possibly important uncertainty or variability Probably no important uncertainty or variability No important uncertainty or variability

26 JUDGEMENT IMPLICATIONS BALANCE OF EFFECTS Favors the comparison Probably favors the comparison Does not favor either the intervention or the comparison Probably favors the intervention Favors the intervention Varies Don't know RESOURCES REQUIRED CERTAINTY OF EVIDENCE OF REQUIRED RESOURCES COST EFFECTIVENESS EQUITY Large costs Moderate costs Negligible costs and savings Moderate savings Very low Low Moderate High Favors the comparison Reduced Probably favors the comparison Probably reduced Does not favor either the intervention or the comparison Probably no impact Probably favors the intervention Probably increased Large savings Varies Don't know Favors the intervention Varies No included studies No included studies Increased Varies Don't know ACCEPTABILITY No Probably no Probably yes Yes Varies Don't know FEASIBILITY No Probably no Probably yes Yes Varies Don't know

27 Conclusions Should LMWH vs. UFH be used in acutely ill medical patients? TYPE OF RECOMMENDATION RECOMMENDATION JUSTIFICATION SUBGROUP CONSIDERATIONS IMPLEMENTATION CONSIDERATIONS MONITORING AND EVALUATION Strong recommendation against the intervention Conditional recommendation against the intervention Conditional recommendation for either the intervention or the comparison Conditional recommendation for the intervention Strong recommendation for the intervention The ASH guideline panel suggests using LMWH, Fondaparinux or UFH rather than no heparin in acutely ill medical patients (conditional recommendation, low certainty of the evidence about effects). However, the ASH guideline panel suggests using LMWH (low certainty of the evidence about effects) or Fondaparinux (very low certainty of the evidence about effects) rather than UFH in acutely ill medical patients (conditional recommendation). This recommendation applies to patients with stroke who receive VTE prophylaxis. This includes stroke patients despite the statistically significant interaction for bleeding risk between non-stroke and stroke patients. The bleeding risk is slightly higher in stroke patients but the panel thought that the desirable consequences of LMWH outweighed the undesirable consequence of UFH. RESEARCH PRIORITIES

28 Appendices Appendix 1 Outcomes Mortality assessed with: all-cause mortality Pulmonary Embolism - representing the moderate marker state assessed with: any PE Proximal Deep Vein Thrombosis - representing the moderate marker state assessed with: any symptomatic DVT of participants (studies) Follow up 8041 (9 RCTs) 8037 (11 RCTs) 3581 (2 RCTs) Quality of the evidence (GRADE) MODERATE a LOW b,c LOW b,c Relative effect (95% CI) RR 0.99 (0.82 to 1.19) RR 0.82 (0.40 to 1.68) RR 0.80 (0.21 to 2.96) Anticipated absolute effects * (95% CI) Risk with UFH Study population Risk difference with LMWH 50 per 1,000 1 fewer per 1,000 (9 fewer to 10 more) Study population 6 per 1,000 1 fewer per 1,000 (4 fewer to 4 more) Low 0 per 1,000 d 0 fewer per 1,000 (0 fewer to 0 fewer) Study population 3 per 1,000 1 fewer per 1,000 (2 fewer to 5 more) Low 2 per 1,000 d 0 fewer per 1,000 (1 fewer to 4 more) Study population

29 Outcomes of participants (studies) Follow up Quality of the evidence (GRADE) Relative effect (95% CI) Anticipated absolute effects * (95% CI) Risk with UFH Risk difference with LMWH 3 per 1,000 1 fewer per 1,000 (2 fewer to 5 more) Distal Deep Vein Thrombosis - representing the moderate marker state assessed with: any symptomatic DVT 3581 (2 RCTs) Major Bleeding 8470 (11 RCTs) LOW b,c LOW b,c RR 0.80 (0.21 to 2.96) RR 0.80 (0.48 to 1.31) Low 1 per 1,000 0 fewer per 1,000 (1 fewer to 2 more) Moderate 14 per 1,000 3 fewer per 1,000 (11 fewer to 28 more) Study population Gastrointestinal Bleeding - not reported Thrombocytopenia 3673 (3 RCTs) MODERATE c RR 0.36 (0.07 to 1.81) 10 per 1,000 2 fewer per 1,000 (5 fewer to 3 more) Study population 3 per 1,000 2 fewer per 1,000 (3 fewer to 2 more) a. Wide confidence interval, including both appreciable benefit, and no benefit, considering the confidence interval around the relative effect and absolute effect. b. Downgraded for risk of bias due to unclear random sequence generation and allocation concealment across majority of trials, and lack of blinding of participants and study personnel as well as unclear blinding of outcome assessors. c. Wide confidence interval, crossing line of no effect, and including both benefit and harm. d. Data from Wang Indirect observational data from a large cohort evaluating thromboprophylaxis among 11,135 patients, 3390 (30.44%) were treated with LMWH, 1629 (14.63%) with warfarin, 3531 (31.71%) with UFH, and 42 (0.38%) with fondaparinux

30 Appendix 2

31 Appendix 3

32 Question 3 Should fondaparinux vs. other pharmacologic agent be used for acutely ill medical patients? POPULATION: acutely ill medical patients BACKGROUND: Hospitalization is a major risk factor for occurrence of thrombosis in acutely ill medical patients. Studies have shown an increased risk of INTERVENTION: fondaparinux venous thromboembolism in the first 30 days after patients discharge. Venous thromboembolism is one of the most significant COMPARISON: other pharmacologic agent preventable causes of morbidity and mortality in hospitalized patients. MAIN OUTCOMES: SETTING: PERSPECTIVE: Assessment PROBLEM Mortality; Pulmonary Embolism - representing the moderate marker state; Proximal Deep Vein Thrombosis - representing the moderate marker state; Distal Deep Vein Thrombosis - representing the moderate marker state; Major bleeding; Inpatient Clinical recommendation - Population perspective Is the problem a priority? No Probably no Probably yes Yes JUDGEMENT RESEARCH EVIDENCE Hospitalization represents a major risk factor for deep vein thrombosis and pulmonary embolism; up to 22% of venous thromboembolisms occur following non-surgical hospital admissions (Heit 2002). Symptomatic venous thromboembolism may complicate over 4.5 per 1000 admissions to the medical services (Zakai 2013), leading to a higher risk of morbidity and mortality. ADDITIONAL CONSIDERATIONS DESIRABL E How substantial are the desirable anticipated effects? Trivial Small See Appendix 1

33 Moderate Large UNDESIRABLE EFFECTS CERTAINTY OF EVIDENCE VALUES How substantial are the undesirable anticipated effects? Large Moderate Small Trivial What is the overall certainty of the evidence of effects? Very low Low Moderate High No included studies Is there important uncertainty about or variability in how much people value the main outcomes? Important uncertainty or variability Possibly important uncertainty or variability Probably no important uncertainty or variability No important uncertainty or variability The relative importance of the outcomes* was as follows in the identified studies: Pulmonary embolism: (Hogg 2013, Hogg 2014, Locadia 2004) Deep vein thrombosis: (Hogg 2013, Hogg 2014, Locadia 2004, Marvig 2015, Utne 2016) Deep vein thrombosis patients' own current health: 0.95 (Time trade off) (Locadia 2004) Panel noted possibly fewer cases of HIT with fondaparinux compared to LMWH and UFH. Quality of evidence downgraded due to indirectness of the comparison and imprecision. Gastrointestinal tract bleeding event: 0.65 (standard gamble and time trade off) (Hogg 2013, Locadia 2004) Muscular bleeding: 0.76 (time trade off) (Locadia 2004)

34 BALANCE OF EFFECTS Does the balance between desirable and undesirable effects favor the intervention or the comparison? Favors the comparison Probably favors the comparison Does not favor either the intervention or the comparison Probably favors the intervention Favors the intervention Minor intracranial bleeding event: 0.75 (standard gamble) (Hogg 2013) Major intracranial bleeding event: 0.15 (standard gamble) (Hogg 2013) Central nervous system bleeding: (standard gamble) (Lenert 1997, O'Meara 1994) Treatment with LMWH: (time trade off) (Marchetti 2001) * indicated by utility value where 0 = death and 1.0 = full health Studies described the following regarding the relative importance of outcomes and patients' preferences for VTE prophylaxis; Patients highly value the benefits of VTE risk reduction of VTE prophylaxis (Haac 2016, Locadia 2004, Quante 2012, Wong 2015) and that they would like to avoid adverse events but most of them are not afraid of the adverse events (Barcellona 2000, Haac 2016, O Meara 1994, Quante 2012, Wong 2015). Patients highly value the benefits of VTE risk reduction of VTE prophylaxis; patients would like to avoid adverse events but most of them are not afraid of the adverse events. Studies additionally described the following regarding the relative importance of outcomes and patients preferences for the pharmacological prophylaxis: Most patients (78%) receiving low molecular weight heparin would like to continue with the same methods (Maxwell 2002).

35 RESOURCES REQUIRED How large are the resource requirements (costs)? Large costs Moderate costs Negligible costs and savings Moderate savings Large savings Cost of disease * See Appendix 2 *: Unless otherwise specified, the costs of disease are based on a systematic review on the USA setting. (Grosse 2016) Treatment of an acute VTE on average appears to be associated with incremental direct medical costs of $12,000 to $15,000 (2014 US dollars) among first-year survivors, controlling for risk factors. Subsequent complications are conservatively estimated to increase cumulative costs to $18,000 23,000 per incident case. According to the analysis of Caprini et al. based on Medicare fee schedules, the predicted first-year treatment costs are $1100 for mild-moderate PTS (without leg ulcers) and $5000 for severe PTS with open ulcers. Cost estimates per bleed event vary widely. The average cost of heparin-induced thrombocytopenia (HIT) is $ Patients who develop CTEPH are much more expensive to treat than other PE patients, with average monthly expenditures of $5500 and $600, respectively. Cost of interventions See Appendix 3 Footnotes: ASP Drug Pricing Files: 2. National Average Drug Acquisition Cost, and The Affordable Care Act Federal Upper Limits (FUL): 3. Good Px: Panel considered the drug price for fondaparinux (generic) to be about $20 per syringe. 4. NHS Electronic Drug Tariff, June ( 1=$1.675) 5. Pounds in 2013 ( 1=$1.550) 6. Medicare CPT (Current Procedural Terminology) Reimbursement for IVC Filter Placement:

36 We also searched other sources for cost information, including Medicare Drug Average Sales Price, National Average Drug Acquisition Cost, The Affordable Care Act Federal Upper Limits (FUL) and GoodPharmacy ( CERTAINTY OF EVIDENCE OF REQUIRED RESOURCES COST EFFECTIVENESS What is the certainty of the evidence of resource requirements (costs)? Very low Low Moderate High No included studies Does the cost-effectiveness of the intervention favor the intervention or the comparison? Favors the comparison Probably favors the comparison Does not favor either the intervention or the comparison Probably favors the intervention Favors the intervention Only for medication costs, warfarin, heparin, aspirin are not costly. Direct oral anticogulants including dabigatran (1 dose pack/60 capsules of dabigatran 150mg: $356 to $472), rivaroxaban (20 mg once daily with for a month: $366 to $496), and apixaban (60 tablets of apixaban 5mg: $366 to $497) cost more. Low molecular weight heparins also increase cost, with enoxaparin (4 syringes 40mg/0.4ml of enoxaparin: $28.72 to $284.00), cost similar to heparin, and higher medication cost for dalteparin (10 syringes of dalteparin 18000iu/0.72ml: $1, to $1,576.44) and fondaparinux (10 syringes of fondaparinux 7.5mg/0.6ml: $4, to $11,320.18). No evidence available for cost-effectiveness.

37 No included studies EQUITY ACCEPTABILITY What would be the impact on health equity? Reduced Probably reduced Probably no impact Probably increased Increased Is the intervention acceptable to key stakeholders? No Probably no Probably yes Yes No research evidence was identified. Acceptability and use of pharmacological prophylaxis: Studies and surveys suggest the following regarding barriers associated with the intervention/option and its use: A survey among 568 physicians and 825 patients from 5 countries showed that more patients considered injectable treatments effective than considered oral treatments effective (87% versus 76%, respectively). This trend was well predicted by the physicians (98% and 61%, respectively). Additionally, 46% of patients would accept an injectable treatment program lasting >2 months (67% for life-threatening diseases), a figure underestimated by physicians (11% and 46%, respectively). Overall, 73% of patients stated they would never miss an injection, where as 54% of physicians expected patients to miss one injection in a month of therapy. (Cimminiello 2012) Among 250 hospitalized (surgical and medical) patients, initiation of prescribed therapy was 95% for LMWH, 88% for UFH 3/day and 87% for UFH 2/day. All scheduled doses were received by 77% on LMWH, 54% on UFH 3/day and 45% on UFH 2/day. Patient refusal explained 39% of omitted LMWH and 44% of omitted UFH doses. LMWH was less likely to be administered in surgical than in medical patients. (Fanikos 2010) A survey among 1,553 Canadian health care providers showed that DVT prophylaxis was perceived as important by all provider groups, but this did